Finite Element Analysis - Buckling Stability

Hello all designers,

I am doing an analysis of buckling stability of a transport barge transverse bulkhead...

simple hand calculations considering the bulkhead and compressive in-plane loads and a lateral pressure yields a high factor against buckling [lambda]..

feel i'm being too simplistic and hence the need for deeper analysis

I have modeled only a part of the barge..the relevant part [in my view] being the bulkhead in question and a length equal to the distance between bulkheads on either ends...my question is about the boundary condition and loads...I want to apply the bending moment and shear forces on the model..but since I do not have the full model, I was a bit confused how to apply them to the partial model...

I have the values of BM and shear forces from the calculation for the full hull..how to I induce the same values of BM and shear forces on the part model ??

and any thoughts on how to apply hydrodynamic loads for the partial hull ??

 

I think a barge, with transverse structure, should have no problems with buckling.
Transverse bulkheads is very difficult to have an appreciable load on the same plane, so it should not be necessary to study buckling.
Furthermore, a calculation using formulas Classification Society, for this type of boat is more than enough. Not be necessary to use FEM analysis unless the case of very large and strange loads.
The contour elements are what you decide, no one else can know, but normal is to consider bulkheads perimeter fixed.
Hull girder must be studied in its entirety. It is little approximate to use just a section between two bulkheads. To reduce the number of nodes, if you want, studying only one side, port or starboard.

 

Thanks for the reply

Hello tansl...in fact there are some heavy loads on deck with high VCG's and hence high loads when there is roll motion....in the FE model, without considering gloabal bending moments, shear stress or any hydrodynamic loads or any lateral loads by the ballast water in the tanks, the factor against bucking is 2.7..that is considering the ends of the modeled geometry as fixed...hence if these loads include could further bring down the factor against buckling...just want to be sure for the client

 

and i am infact considering 1 section on either side of the bulkhead...so 2 sections in total...

 

I always do that loads on deck are supported only by decks's reinforcements (beams and girders). These can be attached at their ends to bulkheads so only part of the total load will be transmitted to the bulkhead. But it is also true that each designer raises his structure in a way.
It is desirable to raise, for bulkheads, the worst possible situations. Is likely to consider only the compressive stress is the worst but there may be others worse if, in addition, you consider the hydrostatic loads, eg one side of the bulkhead.
I think you should expand the section to study the longitudinal strength of the ship-beam or at least change it to a section between two bulkheads.

 

Considering a worst case scenario: does the cargo apply a force against the bulkhead? For example, is this a bulk carrier?

 

Is that hard to understand?

 

It looks to me like you have already included the most important direct loads on the barge bulkhead. It's not clear that adding the shear force and bending moment to the ends of your model will change the loads in the transverse bulkhead very much. These loads, however, should be considered if you think that either the side shell or the deck or bottom structure of the barge might also buckle, leaving the transverse bulkhead boundary unsupported.

There are a lot of details about the barge and its loading which we don't know, so we're unable to give you the full benefit of our professional experience in judging if you need to go further with this analysis.

The ABS has a guide and a commentary publication for performing buckling analyses here:

http://ww2.eagle.org/content/eagle/...ucklingandultstrengthassessoffshorestructures

 

So..??

Remember it is not just about passing an arbitrary rule, it is about fit-for-purpose. In addition, once the structure weight has been calculated, based upon the values of the scantlings you have calculated, is the structure weight within the weight estimate?

There are plenty of guidelines on partial FE models and boundary conditions in the Class rules.

DNV, for example:

"..The model may be supported in the vertical direction by applying vertical springs at the lines formed by the intersections between side and transverse bulkheads, inner side and transverse bulkheads, and longitudinal bulkhead and transverse bulkheads. The spring constant, Ki, should be equally distributed along the line forming the intersection. The spring constant may be calculated as follows, ignoring the effect of bending deflection:.."

Ki = 8.Asi.E/7.8x3.Lt

Asi = shear area of i
Lt = length of one cargo tank
i = side, inner side or long.t BHD.

Again, so what?...if the structure weight on your calculations satisfies the weight estimate, one can consider the structure to be fit-for-purpose.

I never design down to the bare minimum, when doing structures and vessels such as these; that being to Class and a commercial 'workboat' of sorts. Longevity is far more important and less warranty bills too!

 

@TANSL... thanks a lot for your guidance on this matter...i shall definitely expand the section to include 2 bulkheads !!

 

just the lateral force on the bulkhead from the ballast water..I think this should affect the buckling safety factor...

 

thank you for pointing me in the right direction...the ABS guide should be useful in doing the analysys....

 

and a general question for all all..do you all use "DNV-RP-C208 'Determination of structural capacity by non-linear FE-methods' " or a similar guideline when using FEM ??

 

Here extracts:

You simply pro-rate the values along the length as required at the location. Since the moment can be easily changed into a force x lever. The lever being the distance required on your hull and thus the force also varies.

No. I use DNV-RP-C203...which is all you really need. Not necessary for non-linear either.